Concentricity is a geometric tolerance used in engineering drawings to ensure that the central axis of one cylindrical feature aligns with the central axis of another. In the formal language of Geometric Dimensioning and Tolerancing (GD&T), it has a surprisingly specific and often misunderstood definition: concentricity requires that the midpoints of diametrically opposed surface points all fall within a cylindrical tolerance zone centered on a datum axis. Its symbol is two concentric circles, and despite its intuitive-sounding name, it works differently than most people expect.
How Concentricity Actually Works
In everyday language, “concentric” means sharing the same center. Two circles drawn one inside the other are concentric. But in GD&T, the concentricity callout doesn’t simply check whether two features share a center. Instead, it evaluates something called the derived median axis of a part.
Here’s the process. At any given cross-section of a cylindrical feature, you take two points that sit exactly 180 degrees apart on the surface. You find the midpoint between those two points. You repeat this across many pairs of opposing surface points on that cross-section, then move along the length of the cylinder and do it again at additional cross-sections. When you connect all of those midpoints, you get the median axis of the feature. For the part to pass a concentricity check, every one of those median points must fall inside a cylindrical tolerance zone that is centered on the datum axis.
This is a critical distinction. Concentricity doesn’t evaluate the actual surface of the part. It evaluates a derived set of points calculated from surface measurements. That’s what makes it both powerful in theory and difficult in practice.
Why Concentricity Is So Hard to Measure
Measuring concentricity requires mapping median points across the entire feature. At each cross-section, you need a minimum of three pairs of diametrically opposed points to derive the median. Those median points must then be compared against the datum axis to see if they all fall within the specified tolerance zone.
In practice, this usually requires a Coordinate Measuring Machine (CMM) capable of capturing many precise surface points and computing the midpoints mathematically. A simple dial indicator spun around the part surface won’t give you a true concentricity reading, because that measures something different: runout.
Machinists often use runout as a quick sanity check for concentricity, since runout inherently captures some of the same geometric errors. But the two values don’t have a direct mathematical relationship. Runout compares the most extreme surface variations detected across an entire rotation, regardless of where on the circle they occur. Concentricity, by contrast, cares only about the midpoints of points sitting exactly opposite each other. A part can pass a runout check and fail concentricity, or vice versa.
Concentricity vs. Runout vs. Position
These three tolerances all deal with how well cylindrical features line up, but they control different things:
- Concentricity controls the derived median points of a feature relative to a datum axis. It’s the most abstract of the three because it requires calculating midpoints rather than measuring surfaces directly.
- Runout (total runout) is measured by rotating the part 360 degrees around the datum axis while a dial indicator rides on the surface. The runout value equals the maximum variation the indicator detects. Runout inherently controls both circularity and alignment with the datum axis, making it a practical all-in-one check.
- Position controls the location of a feature’s axis (or center) relative to a datum, which is what most engineers actually need when they think they want concentricity. It’s simpler to measure and more commonly understood.
The confusion between these tolerances is one of the main reasons concentricity has a complicated reputation. Many engineers specify concentricity when what they really need is a position tolerance controlling axis-to-axis alignment, or a runout tolerance controlling surface variation.
What Concentricity Doesn’t Guarantee
One of the biggest misconceptions is that concentricity ensures two circles or cylinders are truly concentric in the intuitive sense. It doesn’t. Because concentricity evaluates derived median points rather than actual surfaces, it doesn’t guarantee that the surfaces are round. A feature could have an irregular, slightly out-of-round cross-section and still pass a concentricity check, as long as the midpoints of opposing surface points land within the tolerance zone.
It also doesn’t directly control how the surface behaves during rotation. If your concern is vibration or wobble in a rotating assembly, runout is typically a better choice because it captures surface variation directly.
Removed From the Latest Standard
Concentricity (along with its counterpart, symmetry) was removed from the ASME Y14.5-2018 standard, the governing document for GD&T in the United States. The symbols had been present since the 1982 version of the standard, but they were always considered controversial and complicated.
The core issue was that controlling opposing median points is rarely a true functional requirement. In most real-world applications, what matters is how the axis of one feature relates to the axis of another, or how the surface behaves during rotation. Position tolerancing and runout handle those needs more clearly. Engineers frequently confused the median-point-based definition of concentricity with a simpler axis-to-axis requirement, leading to parts being inspected incorrectly or tolerances being applied that didn’t reflect what was actually needed.
The removal doesn’t mean the concept is gone forever. You’ll still encounter concentricity callouts on older drawings, and in the ISO standard system it lives on under the name “coaxiality.” But for new designs following ASME Y14.5-2018, position and runout are the recommended replacements.
When Concentricity Was the Right Choice
Before its removal, concentricity was technically the correct callout in one narrow scenario: when the functional requirement of the part genuinely depended on mass distribution being balanced around a central axis. Think of a high-speed rotating shaft where even slight asymmetry in material distribution causes vibration. In that case, the median axis (which reflects where the “middle” of the material actually sits) matters more than the surface profile alone.
In practice, even these applications were often better served by runout tolerances combined with balancing operations after machining. The measurement difficulty and frequent misapplication of concentricity meant that most shops struggled to verify it reliably, making it a tolerance that looked good on paper but caused headaches on the shop floor.